Most people open the endpoint catalog and stall in the same place: 46 endpoints, similar names, no obvious starting point. This guide reorganizes them by the job you are doing rather than by endpoint name, so you can find the two or three you actually need.
The full list lives in the API catalog; this page is about choosing.
Start from the question, not the endpoint
| Your question | Group | Count |
|---|---|---|
| Is this category or niche worth entering | Market analytics | 14 |
| Which specific products are making money | Product data | 12 |
| What are shoppers actually searching | Keyword data | 11 |
| Where does a competitor's traffic come from | Traffic analytics | 3 |
| Can I use this brand name | Global trademarks | 4 |
| I need the text out of an image | Image text | 1 |
All 46 endpoints share one authentication scheme and one balance. You do not subscribe per group.
Product research: category first, or product first
There are two entry points, and they run in opposite directions.
Top-down, starting from market size. Market Research filters niches by category node and average monthly units, then Market Statistics, Price Distribution, and Brand Concentration tell you whether a handful of brands already own it.
Bottom-up, starting from a keyword or a product. Product Research filters by keyword, price, rating, seller count, and listing attributes:
curl --request POST \
--url https://ecommercedataapi.com/v1/amazon/product/research \
--header "Content-Type: application/json" \
--header "X-API-Key: ${ECOMMERCE_DATA_API_KEY}" \
--data '{
"marketplace": "US",
"month": "nearly",
"keyword": "bath rug",
"matchType": 2,
"minPrice": 20,
"maxPrice": 80
}'Once you have candidate ASINs there are two ways to estimate sales: BSR Sales Estimate needs only a BSR and a category ID, which suits bulk screening, while Product Sales Estimate works per ASIN. Both return estimates, not platform-reported sales.
Keywords: search terms, traffic terms, and ABA are three different things
They get used interchangeably, but they map to different endpoints:
- Search term discovery — Keyword Miner and Keyword Research expand a seed term into related queries with demand and competition metrics.
- Reverse traffic lookup — Reverse Keyword Lookup takes an ASIN and returns the terms that listing actually gets impressions on. This is the most direct competitor-analysis endpoint.
- ABA rankings — ABA Weekly and ABA Monthly cover search-term rank data, sliced by week or month.
Two more angles: Keyword Expansion works backwards from a set of ASINs, and Google Keyword Trends adds off-platform demand.
Operations monitoring: turning one query into a daily job
After launch you care about different fields. Three endpoints make up the minimum monitoring set:
- Product Details — current price, BSR, rating, review count, seller count, fulfillment.
- Coupon Trends — a competitor's promotion cadence.
- Product History — long-run price and rank curves.
Add Product Competitors and Traffic Sources for the competitive picture. For how to schedule it, see Competitor price and coupon monitoring.
The brand check before you list
The four trademark endpoints get overlooked, but they are the cheapest pre-launch check available: Trademark Search for whether a name is taken, Trademark Details for status and classes, Trademark Statistics for how a term is registered across offices, and Data Coverage for which offices are supported. Worked through in Checking trademarks before you list.
Two ways to call, pick one
REST suits scheduled jobs, warehouses, and product features. Authenticate with the X-API-Key header; the secret is shown once at creation. See authentication.
An AI agent suits ad hoc analysis and anyone who would rather not write code. Install the Skill once or connect the remote MCP server, then describe the task in plain language — no endpoint paths to memorize. See Querying Amazon data in Claude Code without writing code.
Both paths share the same catalog and the same account balance.
Coverage and limits
- 13 marketplaces: US, UK, DE, FR, IT, ES, MX, CA, AU, JP, IN, AE, BR.
- Sales, traffic, and conversion figures are estimates that get revised over time. Re-check with current data before a decision that matters.
- Each successful billable request consumes one API call by default. Rate and concurrency limits are enforced per account, so extra keys do not raise your quota.
- Handling for 429 and the other error codes is in errors and rate limits.